Anal. Chem. 2005, 77, 4176-4184
Application of Ion Trap-MS with H/D Exchange and
QqTOF-MS in the Identification of Microbial
Degradates of Trimethoprim in Nitrifying Activated
Sludge
Peter Eichhorn,*,† P. Lee Ferguson,‡ Sandra Pérez,† and Diana S. Aga†
Chemistry Department, The State University of New York at Buffalo, 611 Natural Sciences Complex,
Buffalo, New York 14260, and Department of Chemistry and Biochemistry, University of South Carolina,
631 Sumter Street, Columbia, South Carolina 29208
In this work, the identification of two microbial degradation products of the antimicrobial trimethoprim (290 Da)
is described. The structural elucidation of the metabolites,
which were produced by nitrifying activated sludge bacteria in a small-scale laboratory batch reactor, was accomplished by electrospray ionization-ion trap mass
spectrometry conducting consecutive fragmentation steps
(MSn) combined with H/D-exchange experiments. Although one metabolite corresponded to r-hydroxytrimethoprim (306 Da), oxidation of the aromatic ring within
the diaminopyrimidine substructure was determined for
the second degradate (324 Da). Accurate mass measurements of the two metabolites were provided by a hybrid
quadrupole time-of-flight-mass spectrometer operated in
MS/MS mode. With absolute mass errors of <5 mDa, it
allowed us to confirm the proposed elemental composition
for the protonated precursor ions as well as for a series
of fragment ions that were previously identified by ion trap
mass spectrometry. The study emphasized the potential
of nitrifying activated sludge bacteria for breaking down
an environmentally relevant pharmaceutical that is otherwise poorly degradable by a bacterial community encountered in conventional activated sludge.
The widespread occurrence of persistent residues of pharmacologically active compounds in natural aquatic environments has
received considerable attention in the past few years.1-3 Since most
pharmaceuticals used in human medicine are eventually disposed
of with the sewage, wastewater treatment facilities play a key role
in removing drugs from the water stream, thus preventing them
from reaching the receiving water bodies. Those compounds that
have a low tendency to partition onto biosolids and are not
amenable to microbial degradation during biological treatment
using activated sludge possess the greatest potential of surviving
* Correponding author. Phone: 1-716-645-6800, x2206. Fax: 1-716-645-6963.
E-mail: [email protected].
†
The State University of New York at Buffalo.
‡
University of South Carolina.
(1) Heberer, T. Toxicol. Lett 2002, 131, 5-17.
(2) Ternes, T. A.; Joss, A.; Siegrist, H. Environ. Sci. Technol. 2004, 38, 392A399A.
(3) Ternes, T. A. TrAC, Trends Anal. Chem. 2001, 20, 419-434.
4176 Analytical Chemistry, Vol. 77, No. 13, July 1, 2005
the treatment plant. Among the frequently detected drugs in
surface waters is trimethoprim,4-6 an antimicrobial compound of
high water solubility that is commonly prescribed in combination
with the sulfonamide sulfamethoxazole for the treatment of
infectious diseases in humans. Regarding the fate of trimethoprim
during sewage treatment, analytical measurements on composite
samples of primary and secondary effluents from two sewage
treatment plants in Switzerland showed that the antimicrobial was
eliminated only to a small extent during the conventional activated
sludge process. While in one plant the trimethoprim concentration
dropped by 24%, no measurable removal was reported for a second
facility studied.7 The lacking capability of the bacterial consortium
in activated sludge to metabolize trimethoprim was corroborated
in laboratory settings. In experimental setups designed to simulate
the biological wastewater treatment, trimethoprim exhibited a
strong resistance to microbial breakdown by activated sludge
bacteria, even after a prolonged adaptation phase of several
weeks.8,9
In contrast to this recalcitrance in typical wastewater treatment
plants, a rapid primary degradation of trimethoprim was achieved
in nitrifying activated sludge collected from a municipal sewage
treatment plant employing a two-stage process composed of an
activated sludge coupled to a nitrification process.8 Such distinct
capability of nitrifying bacteria to degrade recalcitrant compounds
has been reported in several studies for a wide array of aromatic
xenobiotics, for example, anisol,10 aniline,11 and naphthalene,12 in
which ammonia-oxidizing bacteria (AOB), producing an ammonia
(4) Hilton, M. J.; Thomas, K. V. J. Chromatogr., A 2003, 1015, 129-141.
(5) Stackelberg, P. E.; Furlong, E. T.; Meyer, M. T.; Zaugg, S. D.; Henderson,
A. K.; Reissman, D. B. Sci. Total Environ. 2004, 329, 99-113.
(6) Vanderford, B. J.; Pearson, R. A.; Rexing, D. J.; Snyder, S. A. Anal. Chem.
2003, 75, 6265-6274.
(7) Goebel, A.; McArdell, C. S.; Suter, M. J. F.; Giger, W. Anal. Chem. 2004,
76, 4756-4764.
(8) Pérez, S.; Eichhorn, P.; Aga, D. S. Environ. Toxicol. Chem. 2005, 24, in
press.
(9) Junker, T.; Knacker, T.; Römbke, J.; Alexy, R.; Kümmerer, K. Proceedings
of the 24th Annual Meeting SETAC in North America, Austin TX, November
8-13, 2003.
(10) Chang, S. W.; Hyman, M. R.; Williamson, K. J. Proceeding of the 5th
International In Situ and On-Site Bioremediation Symposium, San Diego
CA, Apr. 19-22, 1999; pp 131-135.
(11) Keener, W. K.; Arp, D. J. Appl. Environ. Microbiol. 1994, 60, 1914-1920.
(12) Chang, S. W.; Hyman, M. R.; Williamson, K. J. Biodegradation 2003, 13,
373-381.
10.1021/ac050141p CCC: $30.25
© 2005 American Chemical Society
Published on Web 05/10/2005
Figure 1. LC/ESI-MS chromatogram corresponding to sample from batch reactor spiked with trimethoprim at 20 mg/L. Upper mass trace
shows total ion chromatogram (TIC) acquired over mass range of m/z 100-400 in positive ion mode. Extracted ion chromatograms (XIC) are
shown in the lower part of the graph. The peak in the TIC marked with an asterisk was identified as an impurity in the trimethoprim used
(carrying an ethoxy group instead of a methoxy group; m/z 305).
monooxygenase enzyme of broad specificity, were suggested to
co-metabolize the organic substrates. The assumption of AOB
being involved in the biotransformation process of the pollutants
was supported by findings that the degradation was inhibited in
the presence of specific AOB inhibitors and that the observed
metabolites were compatible with reactions catalyzed by ammonia
monooxygenase.
To gain insight into the degradation processes of the antimicrobial trimethoprim that are brought about by bacteria specific
to nitrifying activated sludge, this study aimed at the identification
and characterization of major metabolites of the target compound.
To this end, the structures of biodegradation products, previously
generated in a small-scale laboratory batch reactor, were elucidated by applying multiple-stage fragmentation studies in combination with H/D-exchange experiments using an electrospray
ionization-ion trap mass spectrometer (ESI-IT-MS). In addition,
a quadrupole time-of-flight-mass spectrometer (ESI-QqTOF-MS)
was used to provide accurate mass measurements, allowing
verification of the assigned chemical structures of the metabolites.
EXPERIMENTAL SECTION
Chemical Standards. Trimethoprim (CAS no. 738-70-5) was
purchased from Riedel de Haën (Seelze, Germany). The organic
solvents acetonitrile and methanol were ACS grade (Burdick &
Jackson, Muskegon, MI). Water was prepared with a Nanopure
Diamond water purifier (Barnstead, Dubuque, IA). Deuterium
oxide was obtained from Cambridge Isotopes Laboratories Inc.
(Andover, MA). Acetic acid-d4 was purchased from Aldrich (St.
Louis, MO).
Biodegradation Experiments. An amber 4-L glass bottle was
loaded with 4 L of nitrifying activated sludge, freshly collected
from the nitrification tank of a municipal sewage treatment plant
(Amherst, NY). Bubbling of air through Teflon tubing into the
test medium provided aeration of the system and suspension of
the sludge particulate matter. The biodegradation of trimethoprim
was studied at two test concentrations: (a) at 20 mg/L for the
preparation of sufficient material required for the mass spectrometric characterization of the metabolites and (b) at 20 µg/L for
obtaining a degradation profile at an environmentally relevant
concentration of trimethoprim.
LC/ESI-MS Analysis. The liquid chromatograph used in this
study was an Agilent Series 1100 comprising the modular
components: quaternary pump, a micro vacuum solvent degasser,
and an autosampler with a 100-well tray. Separations were achieved
on a Thermo Hypersil-Keystone BetaBasic-18 100 × 2.1 mm (3
µm) column equipped with a 10 × 2.1 mm guard column of the
same packing material. The mobile phases were (A) water
acidified with 0.3% formic acid and (B) acetonitrile. The gradient
program started from 95 A/5% B. After 1 min, the portion of A
was linearly decreased to 57% within 10.0 min and further to 5%
within 0.1 min. The latter condition was held for 3.9 min. The
initial mobile phase composition was restored within 0.1 min and
maintained for column regeneration for another 5.9 min, resulting
in a total run time of 20 min. The flow rate was 250 µL/min, and
the injection volume was 20 µL. During the first 1.8 min and the
last 5.9 min of each chromatographic run, the LC stream exiting
the analytical column was directed to the waste via a programmable switching valve integrated in the mass spectrometer. The
mass spectrometric analysis was performed on an Agilent Series
1100 SL single-quadrupole instrument equipped with an ESI
source. All acquisitions were performed under positive ionization
mode with a capillary voltage of +4000 V. Nitrogen was used as
nebulizer gas (35 psi) as well as drying gas at a temperature of
350 °C (10 L/min). For quantitative analysis of trimethoprim and
its metabolites, data were acquired in selected ion monitoring
(SIM) mode recording the protonated molecules, [M + H]+, of
trimethoprim (m/z 291), the metabolite M306 (m/z 307), and the
Analytical Chemistry, Vol. 77, No. 13, July 1, 2005
4177
metabolite M324 (m/z 325) using fragmentor values of 140. The
ions m/z 261, 289, and 181, which represented the most intense
fragment ions detected at a fragmentor value of 230, were used
as qualifier ions for trimethoprim, M306 and M324, respectively.
Data acquisition and processing was done with the software
Chemstation Rev. A.09.03.
LC/ESI-IT-MS Analysis. The ThermoFinnigan system used
for multiple-stage MS experiments consisted of the following
components: Surveyor pump, Surveyor autosampler, and LCQ
Advantage ion trap mass spectrometer, equipped with an ESI
interface. The analytical column, mobile phase composition, and
gradient elution were the same as described above for the singlequad MS except for the H/D-exchange experiments, for which
mobile phase A was D2O containing 0.3% CD3COOD. The flow
rate was 200 µL/min; the injection volume was 10 µL. The needle
voltage of the ESI interface was set to +5000 V, the sheath gas
flow was 30, the temperature of the ion transfer tube was 300 °C,
and the capillary voltage was 10 V. For MSn experiments,
precursor ions were selected with an isolation width of 1.5 Da,
except for the analysis of H/D-exchanged analytes, for which the
isolation width was set to 1.0 Da. Relative collision energies were
between 25 and 40%. Helium was used as both damping gas and
collision gas at a pressure of ∼1 mTorr.
LC/ESI-QqTOF-MS Analysis. Accurate mass MS and MS/
MS analysis of trimethoprim metabolites was performed using a
Waters/Micromass Q-TOF API-US system coupled to an Agilent
1100 Nanobore HPLC system. Diluted batch reactor samples were
separated on an Agilent Zorbax 300SB-C18 nanobore column (150
× 0.1 mm, 3.5-µm particles) prior to MS analysis. The mobile
phases were (A) water acidified with 0.1% formic acid and (B)
acetonitrile with 0.1% formic acid. The flow rate was 500 nL/min,
with a gradient beginning at 95% A/5% B for 2 min and increasing
linearly to 5% A/95% B over 10 min. The injection volume was 0.5
µL. The column was directly interfaced to the nanoelectrospray
ion source using a tapered fused-silica electrospray emitter (New
Objective), and the required electrospray voltage (+2800 V) was
applied through the stainless steel column endfitting. External
mass calibration was performed just prior to analysis using Na(NaTFA)n+ clusters from 50 µM sodium trifluoroacetate in 50:50
acetonitrile/water. This calibrant was also used as a source of
reference internal lock mass (m/z ) 430.9142) for accurate mass
measurement of precursor and product ions from the trimethoprim metabolites. The instrument was operated at a resolution of
10 000 (fwhm), and (+)ESI mass spectra were acquired at 1-s
intervals. High-resolution product ion spectra of M306 and M324
metabolites were acquired using argon as a CID gas with collision
energies of 25 eV.
Sample Preparation. Samples collected from the batch
reactor amended with trimethoprim at a concentration of 20 µg/L
were centrifuged for 4 min at 12000g. An aliquot of the supernatant
was injected into the LC/ESI-MS without further treatment. For
structural elucidation of the degradation products employing LC/
ESI-IT-MS and LC/ESI-QqTOF-MS, a 100-mL sample from the
batch reactor spiked at 20 mg/L was preconcentrated by solidphase extraction (SPE). The sample was first filtered (1.2-µm glass
fiber) and then loaded onto a 6-mL/500-mg OASIS hydrophiliclipophilic (HLB) SPE cartridge (Waters, Milford, MA) coupled in
tandem to a 6-mL/500-mg Sep-Pak tC18 SPE cartridge (Waters).
4178 Analytical Chemistry, Vol. 77, No. 13, July 1, 2005
Figure 2. Degradation profile of trimethoprim in batch reactor.
Nitrifying activated sludge was amended with 20 µg/L of the test
compound.
This combination of two different sorbent materials was used to
ensure high recoveries of the metabolites. The dried cartridges
were eluted separately with 3 × 2 mL of methanol, and the
combined eluates were evaporated to dryness under a gentle
stream of nitrogen. The residue was reconstituted in 1 mL of 10%
acetonitrile/90% water. For the H/D-exchange study, a second
batch-reactor sample was processed in the same way as described
above using SPE. The dried residue, however, was taken up in 1
mL of 10% acetonitrile/90% D2O which contained 0.3% CD3COOD
as catalyst for the H/D exchange.
Quantification. To determine the concentration of trimethoprim and to estimate those of its two metabolites in the samples
from the batch reactor spiked at 20 µg/L, external calibration with
trimethoprim as reference substance was performed for all three
compounds over a linear range from 0.02 to 25 µg L-1, with the
assumption that they have identical ionization efficiencies and
response factors. In view of the relatively narrow retention time
range and the structural similarities enabling protonation at the
amino group, this approach provided a reasonable estimate for
the metabolite levels. As far as matrix ionization suppression or
enhancement effects were concerned, this issue had been addressed in a previous work of our group comparing possible
matrix effects in different types of waters analyzed on the Agilent
LC/ESI-MS.8 Preparation of the calibration series in distilled water
versus not-concentrated sewage had shown no significant impact
on the relative signal intensity for trimethoprim and a series of
sulfonamide antibiotics.
RESULTS AND DISCUSSION
Degradation of Trimethoprim. To investigate the microbial
degradation of trimethoprim in a nitrification tank, a batch reactor
loaded with freshly collected nitrifying activated sludge was spiked
with the parent compound at a concentration of 20 mg/L. Such
high a concentration was chosen to allow the detection of novel
metabolites in solution. Screening of the reactor samples by LC/
ESI-MS over a scan range from m/z 100 to 400 showed the
emergence of two peaks at 8.6 and 8.9 min as the signal of
Figure 3. (a) (+)-ESI-MS2 spectrum of protonated trimethoprim, [M + H]+ ) m/z 291; and (b) (+)-ESI-MS2 spectrum of trimethoprim after H/D
exchange, [M(D4) + D]+ ) m/z 296.
trimethoprim (10.0 min) disappeared (Figure 1). These two peaks,
which were not observed in the control batch reactor loaded with
sterilized sewage, had molecular ions, [M + H]+, of m/z 307 and
325 (metabolites termed as M306 and M324). On the basis of the
order of elution in the reversed-phase column (Figure 1), it was
deduced that the microbial transformation of trimethoprim resulted in more polar products, suggesting an oxidative breakdown
of the antibiotic. A second experiment with a test concentration
of 20 µg/L was carried out to illustrate that the degradation route
was independent of the initial concentration of trimethoprim.
Compound-specific analysis by LC/ESI-MS operated in SIM mode
provided the degradation profile depicted in Figure 2. The instant
decline of the trimethoprim concentration without lag phase was
paralleled by the concurrent formation of the metabolites M306
and M324, suggesting that both compounds were primary
degradation products from the direct transformation of the parent
drug. As can be seen in Figure 2, the transformation rate of
trimethoprim decreased considerably after 2 days, and by day 5,
a steady level was attained with ∼25% of the starting material
remaining undegraded. As to possible elimination of trimethoprim
from the liquid phase through adsorption onto the sludge, previous
studies conducted in identical reactor settings proved this removal
pathway to be insignificant.8 Regarding the two transient metabolites, they built up to a relatively constant concentration of ∼2%
of the initial trimethoprim, indicating that they were subject to
further degradation. Adsorption of the intermediates onto the
biomass was unlikely in view of the increase in polarity relative
to the parent compound. The incomplete primary degradation of
trimethoprimsalso observed in the 20 mg/L experimentswas
attributed to the gradual acidification of the test medium from
the initial of pH 7.2 down to 4.8 by day 5. This was due to the
acid-producing nitrification process, that is, the oxidation of
ammonium to nitrate. Because the optimal pH range for nitrifying
bacteria is between 7.5 and 8.6,13 the considerable decrease of
the pH resulted in a strong reduction of the metabolic activity of
those microorganisms involved in the degradation of trimethoprim. A complete degradation of trimethoprim, however, was
achieved when the pH of the test liquor was maintained in the
slightly alkaline range using ammonium hydroxide for adjusting
the pH (data not shown).
Structure Elucidation. As far as the fragmentation pattern
of the parent drug trimethoprim is concerned, selection of the
protonated precursor ion (m/z 291) in the ion trap and subsequent
activation resulted in the (+)-ESI-MS2 spectrum shown in Figure
3a. All of the fragment ions with m/z values between 230 and 276
were based on a structure comprising the benzene ring, the
bridging methylene group, and the intact diaminopyrimidine. That
the latter moiety did not undergo any fragmentation under the
conditions applied was corroborated by the MS2 spectrum of the
H/D-exchanged trimethoprim ([M(D4) + D]+ at m/z 296), which
contained a total of four exchangeable hydrogen atoms in the
amino groups (Figure 3b). The increment of the ion masses by
five units (m/z’s 230 f 235, 245 f 250, 258 f 263, 261 f 266,
275 f 280, and 276 f 281) indicated that the two deuterated
amino groups (-ND2) were present in all of these fragment ions.
The structures of the ions m/z 230, 261, and 275, as depicted in
(13) Albertson, O. E. Nutrient Control - Manual of Practice FD-7 Facilities
Design: Water Pollution Control Federation:Alexandria, VA, 1983.
Analytical Chemistry, Vol. 77, No. 13, July 1, 2005
4179
Figure 4. (a) (+)-ESI-MS2 spectrum of metabolite M306 (m/z 307); (b) (+)-ESI-MS3 spectrum of metabolite M306 (m/z 307 f 289); and (c)
(+)-ESI-MS3 spectrum of metabolite M306 (m/z 307 f 274). Insets show analogous spectra of M306 after H/D exchange ([M(D5) + D]+ ) m/z
313).
Figure 3a, were consistent with those proposed by Barbarin et
al.14 for the fragmentation pathway of trimethoprim in a triplequadrupole mass spectrometer. The structures suggested for the
three ions with m/z 245, 258, and 276 (Figure 3a) derived from
losses and rearrangements within the trimethoxybenzene portion.
In addition to the assigned product ions encompassing both
aromatic rings, two further fragment ions, m/z 181 and 123, were
observed arising from the cleavage of the protonated molecule at
either side of the central methylene group (C7 atom; see Figure
1 for atom numbering).14 Stabilization of the positive charge could
be readily achieved through mesomeric resonance across the
trimethoxybenzene and diaminopyrimidine rings, respectively
(possible structures shown in Figure 3a).
(14) Barbarin, N.; Henion, J. D.; Wu, Y. J. Chromatogr., A 2002, 970, 141-154.
4180
Analytical Chemistry, Vol. 77, No. 13, July 1, 2005
Of the two metabolites produced in the nitrifying activated
sludge, M306, eluting in the LC chromatogram 1.5 min prior to
trimethoprim (see Figure 1), had a molecular weight of 16 Da
([M + H]+, m/z 307) higher than trimethoprim, suggesting that
the antibiotic had undergone an oxidative transformation; i.e., the
degradate formed was bearing an additional oxygen atom. In
contrast to the signal-rich mass spectrum of the parent compound,
the product ion profile of m/z 307 generated in the IT-MS merely
displayed two fragment ions: a base peak at m/z 289 and a less
intense ion at m/z 274 (Figure 4a). The neutral loss of 18 Da,
presumably the expulsion of water, appeared to be strongly
favored over other fragmentation processes involving cleavages
from the methoxy moieties, as observed for trimethoprim (see
Figure 3a). Confirming evidence that the loss of 18 Da was due
Figure 5. Proposed fragmentation pathway of the protonated M306 under (+)-ESI conditions (ion masses obtained for H/D-exchanged molecule
in parentheses).
to water was provided by the (+)-ESI-MS2 spectrum of M306 after
H/D exchange. First, the deuterium adduct, [M + D]+, was
detected at m/z 313; i.e., the microbial conversion of trimethoprim
had increased the number of exchangeable protons in the
molecule by one, for which a hydroxy group could account.
Second, the MS2 spectrum of m/z 313 (inset in Figure 4a) showed
a fragment ion at m/z 293, which was consistent with the
elimination of D2O (20 Da). To eliminate water from the protonated molecule of the metabolite M306, the C-H bond subject
to oxidation had to be located such that formation of a double
bond was feasible upon expulsion of water. Thus, transformation
of an N-H bond to N-OH could be ruled out, since it would not
alter the number of exchangeable protons. Only the C-H bond
of the methylene group (C7 atom) linking the trimethoxybenzene
to the diaminopyrimidine met this requirement, whereas oxidation
of one of the methoxy groups or of an aromatic C-H bond could
be excluded, since the resulting hydroxy group would lack a
hydrogen on the β-atom essential for formation of a double bond.
The driving force behind the elimination of water during resonance excitation in the ion trap was hypothesized to be formation
of a conjugated system (m/z 289) extending over both rings, as
depicted in Figure 5. As for the fragment ion m/z 274 (Figure
4a) differing from m/z 289 by 15 Da, it was assigned to the loss
of a methyl radical in analogy to the generation of m/z 276 from
the protonated trimethoprim (see MS2 spectrum in Figure 3a).
While the detection of the ion m/z 278 in the spectrum of the
H/D-exchanged M306 (inset in Figure 4a) confirmed that the
neutral fragment expulsed did not carry any exchangeable protons,
the detection of m/z 274 in the MS3 spectrum of m/z 307 f 289
proved that this ion could be formed by m/z 289, that is, by the
precursor ion [M + H - H2O]+ (Figure 4b). The product scan of
m/z 289 did not generate any fragment ions originating from a
central cleavage of the molecule, emphasizing the particular
stability of the aromatic structure of all fragment ions presented
in Figure 5. The structural identification of the two ions with m/z
259 and 243 that were detected in the product ion profile of m/z
289 (Figure 4b) was accomplished by performing an MS3 scan
on the sequence m/z 306 f 274 (Figure 4c). Plausible structures
of the ions m/z 259 and 243 were traced back to the elimination
of a methyl radical and a methoxy radical, respectively. This
interpretation was in agreement with the m/z values of the
fragment ions in the analogous MS3 spectra acquired for the H/Dexchanged M306 (m/z 313 f 278, inset in Figure 4c). The
structures proposed for the ions m/z 259 and 243 in Figure 5 were
also consistent with the ion profile of the MS4 scan on m/z 306 f
289 f 259 (spectrum not shown). The ion m/z 243 was not
detected in this spectrum, because the o-benzoquinone structure
of m/z 259 could not readily give rise to the benzooxirene
structure in the ion m/z 243.
As far as the metabolite M324 is concerned, its molecular
weight was 34 Da higher relative to trimethoprim (molecular
weight ) 290 Da), indicating oxidation or addition of a functional
group. Isolation and activation of the protonated molecule, m/z
325, yielded the (+)-ESI-MS2 spectrum shown in Figure 6a.
Detection of the base peak ion at m/z 181, likewise observed in
the product ion profile of the protonated trimethoprim, indicated
that the trimethoxybenzene portion of the molecule had not
undergone any structural modifications. The strong abundance
of m/z 181, as compared to its intensity in the trimethoprim
spectrum, suggested its formation is favored over other fragmentation processes. Upon cleavage of the C5-C7 bond, conservation
of the positive charge in the other part of the molecule resulted
Analytical Chemistry, Vol. 77, No. 13, July 1, 2005
4181
Figure 6. (a) (+)-ESI-MS2 spectrum of protonated M324, [M + H]+ ) m/z 325; (b) (+)-ESI-MS2 spectrum of M324 after H/D exchange,
[M(D6) + D]+ ) m/z 332; (c) M324, (+)-ESI-MS3 of m/z 325 f 308; (d) M324, (+)-ESI-MS3 of m/z 325 f 307; and (e) M324: (+)-ESI-MS3 of
m/z 325 f 290.
in the formation of the fragment ion m/z 143. Unlike the ion m/z
123 of trimethoprim including the heterocyclic ring, the ion m/z
143 did not comprise the methylene group, suggesting that the
C5-atom of the former diaminopyrimidine ring was a quaternary
carbon atom, because in this instance, the loss of the aromatic
structure of this ring would disable the stabilization of the positive
charge, as postulated for m/z 123 (for mesomeric structures see
Figure 3a). The two fragment ions m/z 308 and 307 were
attributed to losses of ammonia and water, respectively, implying
that the metabolite contained both an aliphatic hydroxy group (as
in M306) and an aliphatic amino group (no loss of ammonia was
observed for the aromatic amino groups in trimethoprim and
M306). Unless an additional amino group had been introduced
into the molecule by the action of nitrifying bacteria, the diaminopyrimidine ring lost its aromaticity as just put forward. Evidence
for the presence of the aliphatic -OH and -NH2 groups was
provided by the product ion scan of the H/D-exchanged M324.
On one hand, the deuterated adduct ion, [M + D]+, of this species
was detected at m/z 332 (Figure 6b); i.e., this molecule was
bearing six exchangeable protons, as compared to four in
trimethoprim; whereas one of the two exchangeable protons was
4182
Analytical Chemistry, Vol. 77, No. 13, July 1, 2005
assigned to the hydroxy group, the second could be explained as
a consequence of the loss of aromaticity converting the N1-atom
or N3-atom into a secondary nitrogen atom. On the other hand,
the fragment ion detected at m/z 312 (∆m/z ) 20) was assigned
as the elimination of D2O, ND3, or both. The observation of m/z
313, that is, the neutral loss of 19 Da, indicated that a non-acidic
hydrogen could be abstracted during the fragmentation process,
leading to the expulsion of HOD, HND2, or both. A probable
position of the hydroxy group was at the C5-atom.
A further key fragment in the structural elucidation of M324
was the ion m/z 297 (in the H/D-exchanged spectrum at m/z
304). The mass difference between the precursor and the product
ion of 28 Da in both spectra in Figure 6a and b proved that this
fragment did not contain any exchangeable hydrogen atoms. Of
the four possible elemental compositions, CH2CH2, CH2N, N2, and
CO, the former three appeared to be rather unlikely to originate
from M324. That CO in turn did not originate from a methoxy
substituent in the benzene ring, but from a loss out of the
structurally modified pyrimidine ring, was shown by generating
the (+)-MS3 spectrum of m/z 325 f 297, in which m/z 181
represented a major fragment ion (spectrum not shown). A
Figure 7. Major fragmentation routes of protonated M324 under (+)-ESI conditions as obtained by multiple-stage experiments.
plausible explanation for the expulsion of CO was derived from
the oxidation of the C4-atom because this was the only secondary
carbon atom in the ring. The loss of CO from a lactame (m/z 290
f 262, Figure 6e) was then feasible under ring contraction,
yielding a five-membered ring (Figure 7). Regarding the fragment
ion m/z 290 observed in the MS2 spectrum of [M + H]+ of M324
(Figure 6a), its identity could be determined by acquiring the (+)ESI-MS3 spectra of [M + H - NH3]+ and [M + H - H2O]+, that
is, the sequence m/z 325 f 308 (Figure 6c) and m/z 325 f 307
(Figure 6d), respectively. Since both MS3 spectra revealed the
presence of m/z 290, this ion corresponded to [M + H - H2O NH3]+. The composition of the ion m/z 280 in the first-generation
spectrum of the protonated M324 (Figure 6a) could also be
confidently assigned with the aid of the product ion profile of [M
+ H - NH3]+ (Figure 6c). The (+)-ESI-MS3 spectrum likewise
showed the fragment ion m/z 280; thus, it corresponded to the
ion [M + H - NH3 - CO]+.
The scheme presented in Figure 7 compiles all the details from
the above interpreted fragmentation pathways along with the
structures of the major identified fragment ions of the protonated
M324 (in dotted-lined box). In addition to the mass spectral
elucidation already provided, the following analysis provided
further support for the metabolite identity. Of the high-molecularweight fragment ions (gm/z 280), all but m/z 290 produced the
ion m/z 181 (see Figure 6a-e), originating from the breakage of
the C5-C7 bond (Figure 7). As an explanation, it was put forward
that the ion m/z 290 possessed a highly conjugated (aromatic)
structure similar to the one of m/z 289, which was formed by
dehydration of the protonated M306 (see Figure 5). The mass
spectrum of m/z 290, that is, MS3 on m/z 325 f 290 (Figure 6e),
exhibited one ion at m/z 262 corresponding to the expulsion of
CO under ring contraction, whereas the ions detected at m/z 275
and 259 were believed to involve losses of radicals out of the
trimethoxybenzene moiety. Here, the formation of m/z 275 could
occur in analogy to the production of m/z 274 by m/z 289 (see
Figure 4b and Figure 5). With regard to the fragment m/z 143
observed in the product ion spectrum of the protonated M324
(Figure 6a), cleavage of the bond C5-C7 led to the formation of
a cyclohexenedione, for which a large number of tautomeric
structures (enol/keto as well as imine/enamine) could be described, thus allowing efficient stabilizization of the positive charge.
The formation of the minor ion m/z 221(Figure 6a), in turn, was
favored by the aromaticity of the benzopyrilium core (Figure 7).
To provide ultimate evidence for the identity of M306 and
M324, accurate mass measurements were performed using
QqTOF-MS, selecting the protonated molecules as precursor ions
for (+)-ESI-MS2 experiments. The results of these analyses are
listed in Table 1, along with the calculated values of the postulated
ion masses of the metabolites and their fragment ions, as well as
the absolute mass measurement errors for each species detected.
Due to fundamental differences between collision-induced dissociation, employed in the QqTOF-MS, and resonance excitation,
used in the IT-MS, the MS/MS spectra obtained for the trimethoprim metabolites using these two instruments were not
identical. However, most major fragment ions were observed in
both spectra, and the fragment ion masses determined using the
Analytical Chemistry, Vol. 77, No. 13, July 1, 2005
4183
Table 1. Accurate Mass Measurements of Trimethoprim and Its Two Metabolites, M306 and M324, As Determined
by LC/ESI-QqTOF-MS in MS/MS Modea
elemental
composition
calcd mass (m/z)
meas mass (m/z)
abs. error (mDa)
DBE
C14H19N4O3
Trimethoprim
291.1457
291.1446
-1.1
7.5
[M + H]+
[M + H - H2O]+
[M + H - H2O - •CH3]+
[M + H - H2O - (CH3)2]+
[M + H - H2O - CH3 - OCH3]+
C14H19N4O4
C14H17N4O3
C13H14N4O3
C12H11N4O3
C12H11N4O2
Metabolite M306
307.1406
289.1301
274.1066
259.0831
243.0882
307.1404
289.1313
274.1057
259.0824
243.0872
-0.2
+1.2
-0.9
-0.7
-1.0
7.5
8.5
9.0
9.5
9.5
[M + H]+
[M + H - H2O]+
[M + H - CO]+
[M + H - NH3 - H2O]+
[M + H - NH3 - CO]+
m/z 221
m/z 181
m/z 143
C14H21N4O5
C14H19N4O4
C13H21N4O4
C14H16N3O4
C13H18N3O4
C12H13O4
C10H13O3
C4H7N4O2
Metabolite M324
325.1512
307.1406
297.1563
290.1141
280.1297
221.0814
181.0865
143.0569
325.1522
307.1364
297.1578
290.1118
280.1344
221.0795
181.0855
143.0596
+1.0
-4.2
+1.5
-2.3
+4.7
-1.9
-1.0
2.7
6.5
7.5
5.5
8.5
6.5
6.5
4.5
3.5
molecular ion/fragment ion
[M + H]+
a
DBE: double bond equivalents
QqTOF-MS were determined to be within e5 mDa of the
theoretical values (within instrument specifications for the mass
range of the fragment ions), and mass accuracies of e5 ppm were
achieved for the protonated pseudomolecular ions of trimethoprim,
M306 and M324.
In conclusion, the microbial degradate M306 was unequivocally
identified as R-hydroxytrimethoprim. This compound, along with
metabolites originating from O-demethylation and ring N-oxidation, had been described as one of the major trimethoprim
metabolites formed after administration to mammals, including
dogs, rats, and humans.15,16 It was also determined in in vitro
studies as an oxidative metabolite of trimethoprim using cell
culture media and microsome incubation mixtures.17 The second
metabolite identified in this work, M324, had not been reported
in higher organisms, and thus, its formation is hypothesized to
be strictly confined to the microbial community in nitrifying
activated sludge.
Environmental Relevance. The degradation studies revealed
that nitrifying sludge bacteria were capable of facilitating an
oxidation of trimethoprim, a pharmaceutical which is not amenable
(15) Schwartz, D. E.; Vetter, W.; Englert, G. Arzneim.-Forsch. 1970, 20, 18671871.
(16) Meshi, T.; Sato, Y. Chem. Pharm. Bull. 1972, 20, 2079-2090.
(17) van’t Klooster, G. A.; Kolker, H. J.; Woutersen-van Nijnanten, F. M.;
Noordhoek, J.; van Miert, A. S. J. Chromatogr. 1992, 579, 354-360.
4184
Analytical Chemistry, Vol. 77, No. 13, July 1, 2005
to biological degradation in a conventional activated sludge
process. This represents an important outcome in view of the great
potential of removing other polar persistent drugs from sewage
and ultimately of minimizing the contamination of natural water
systems. Instead of applying cost-intensive tertiary treatment
technologies, such as advanced oxidation processes, the integration of biological treatments making use of nitrifying bacteria,
which flourish in low organic carbon, ammonium-enriched environments, is worth consideration as an alternative strategy.
Further research in this direction is needed.
ACKNOWLEDGMENT
This material is based upon work supported by the National
Science Foundation under Grant No. 0233700. Any opinions,
findings, and conclusions or recommendations expressed in this
material are those of the authors and do not necessarily reflect
the views of the National Science Foundation. This study was also
supported by a grant from the Interdisciplinary Research and
Creative Activities Fund of the University at Buffalo. S. Pérez
acknowledges a postdoctoral fellowship from the Spanish Ministry
of Education, Culture and Science (EX2003-0687).
Received for review January 24, 2005. Accepted April 7,
2005.
AC050141P